July 18, 2005
Photovoltaic Concentrators To Reach Cost Effectiveness In Few Years?

How close are we to cost effective photovoltaic cells?

Golden, Colo. Solar concentrators using highly efficient photovoltaic solar cells will reduce the cost of electricity from sunlight to competitive levels soon, attendees were told at a recent international conference on the subject. Herb Hayden of Arizona Public Service (APS) and Robert McConnell and Martha Symko-Davies of the U.S. Department of Energy's National Renewable Energy Laboratory (NREL) organized the conference held May 1-5 in Scottsdale, Ariz.

"Concentrating solar electric power is on the cusp of delivering on its promise of low-cost, reliable, solar-generated electricity at a cost that is competitive with mainstream electric generation systems," said Vahan Garboushian, president of Amonix, Inc. of Torrance, Calif. "With the advent of multijunction solar cells, PV concentrator power generation at $3 per watt is imminent in the coming few years," he added.

We have seen steady progress in photovoltaic concentrator technology. We are working with advanced multijunction PV cells that are approaching 38% efficiency, and even higher is possible over time. Our goal is to install PV concentrator systems at $3 per watt, which can happen soon at production rates of 10 megawatts per year. Once that happens, higher volumes are readily achieved," Hayden, Solar Program Coordinator at APS, said.

Growth in the photovoltaic (PV) concentrator business was reflected in the conference attendance, three times that of the 2003 version. This rapid growth was attributed to recent PV concentrator installations and sales forecasts along with excitement created by new solar cell efficiencies approaching 40%. At the conference, NREL announced a new record efficiency of 37.9 percent at 10 suns, a measure of concentrated sunlight. Soon thereafter Boeing-Spectrolab, under contract to NREL and the Department of Energy, surpassed the NREL record with 39.0 percent at 236 suns announced at the European photovoltaic conference in Barcelona, Spain. The efficiency of a solar cell is the percentage of the sun's energy the device converts to electricity.

Photovoltaic (PV) concentrator units are much different than the flat photovoltaic modules sold around the world; almost 1,200 megawatts of flat PV modules were sold last year. PV concentrators come in larger module sizes, typically 20 kilowatts to 35 kilowatts each, they track the sun during the day and they are more suitable for large utility installations.

Those 1,2000 megawatts of flat PV modules sold last year are equivalent to 1 nuclear power plant running only part of the day. So maybe they equal a third or a quarter of a nuclear power plant. However, see the following article where one person is quoted estimating 14,000 megawatts of PV sold in the last year in the world.

Note the concentrator installations are more complex because they have mechanical components to keep the photovoltaics pointed at the sun. This is probably not practical for home roof photovoltaics due to materials, installation, and maintenance costs. Then will large commercial photovoltaics electric power generator facilities become cost effective before residential solar power?

Update: The San Francisco Chronicle has an article about growing venture capital funding of photovoltaics start-ups. Venture capital start-ups are pursuing flexible and cheap plastic photovoltaics.

Nanosys and Nanosolar in Palo Alto -- along with Konarka in Lowell, Mass. -- say their research will result in thin rolls of highly efficient light-collecting plastics spread across rooftops or built into building materials.

These rolls, the companies say, will be able to provide energy for prices as low as the electricity currently provided by utilities, which averages $1 per watt.

Note that the $3 per watt hope from the first article is 3 times the $1 watt figure to compete against utilities.

The Sand Hill Road venture capitalists are interested in photovoltaic materials that require far less capital equipment to produce.

"Silicon is very capital-intensive. You don't need a clean room for plastic power where capital costs are one-tenth of silicon," said Raj Atluru, managing director at the venture capitalist firm of Draper Fisher Jurvetson in Menlo Park, a major investor in Konarka.

Cheap solar power is inevitable. But when?

Share |      Randall Parker, 2005 July 18 12:17 PM  Energy Solar

Brock said at July 18, 2005 1:40 PM:

Randall said:

"Note the concentrator installations are more complex because they have mechanical components to keep the photovoltaics pointed at the sun. This is probably not practical for home roof photovoltaics due to materials, installation, and maintenance costs. Then will large commercial photovoltaics electric power generator facilities become cost effective before residential solar power?"

And the answer is: Soon!


"a young Caltech physics grad named Kevin Hickerson figured out how to reduce the number of motors needed to move 25 mirrors independently, a major cost factor. Instead of two motors for each mirror - the traditional approach - Hickerson's solution requires only two motors for any number of mirrors. The key is a mathematical curve known as the conchoid of Nicomedes (named for the ancient Greek mathematician, who discovered it). A grid of ball bearings arrayed to match the conchoid is attached to a frame inside the Sunflower. As the motors move the frame, the bearings control each mirror's position individually.

The resulting Sunflower 250 is heavy enough to stay put in high winds, but light enough to be lifted by two installers. To take full advantage of outsourced manufacturing, it's sized to fit into a shipping container; commercial units could be transported to your favorite big-box retailer's rooftop direct from the Shenzhen factory.


Energy Innovations' figures show that the Sunflower has a 30 percent cost advantage over typical PV panels before rebates and, in most locations, an even bigger advantage after. Consider a hypothetical Los Angeles light-manufacturing business with 35,000 square feet of roof space. A $684,000 investment (after rebates) in PV panels would generate 90 percent of the company's annual power needs and save roughly $52,000 a year on its electric bill. An array of 750 Sunflowers would deliver the same benefits for $228,000.

Gross hopes to push the price down another 20 percent within two years as manufacturing scale and more efficient silicon kick in. His focus is the metric nearest the hearts of penny-counting CFOs and facilities managers who rule all those endless miles of sprawling rooftops: payback period. 'Right now, PV solar has a 20-year payback, but people are still buying it,' he says. 'Our target for California is five. In Phoenix we could do 3.3.'"

Randall Parker said at July 18, 2005 2:01 PM:


Wow, excellent find.

Yes, payback period is key. I love seeing the payback period quoted per location too.

Cheap photovoltaics will give an economic advantage to areas such as Arizona where heavy sunlight falls.

Note how in the worst month of the year much of the American West as far north as Wyoming gets a lot of solar radiation.

Another point: If payback period is about to undergo a big drop then it doesn't make sense to buy now. You can wait a few years and get a much shorter payback period when prices fall.

Also, I'd love to see the payback period figures without tax rebates.

jim moore said at July 18, 2005 3:11 PM:

Dang, Brock beat me to it,
I think that the Sunflower is really interesting technology, but I would have liked to use a Sterling Engine instead of photovoltaics. The big advantage of using a sterling engine is when the sun is not shining you can burn natural gas to create your heat source. If the solar collectors are on top of your house / business you also have the potential to use the waste heat and bump up the efficiency of the system (at least in the winter).

gmoke said at July 18, 2005 3:43 PM:

When will PVs be cost effective? In some cases, they already are. The lights I use for my bike at night are solar - a flashing red LED that is solar powered that cost $15 and a white steady light I clip to my helmet that cost $30. They are both cost effective and competitive within that marketplace. (Funny, I never think about "payback period" when I'm buying a flashlight. It's more like quality, fitness for the use, and price.) I am about to experiment with using a couple of NewLite solar lanterns for my bedroom reading light (a cost of about $100) and my bedside radio has been a solar/dynamo flashlight/radio that also charges AA batteries for the last few years (about $70, including the custom modifications to enable AA charging). The sooner those interested recognize that some cost effective solar tools are already on the market and begin to act on it, ie buy and use the products, the sooner that marketplace will expand and more powerful applications will become cost effective.

BTW, I always liked Charles Greeley Abbot's clockmotor and counterweight solar tracker for his parabolic trough cooker on Mt Palomar back in the early 20th century and wonder whether Steve Baer is still making his passive trackers using a temperature sensitive working fluid (originally it was Freon) at Zomeworks.

Tom said at July 18, 2005 6:04 PM:

Hmm...I'm paying 6.2 cents a KWH. So that's 54 cents/watt/year.

Nanosys thinks they might get it down to $1/watt. I'm guessing that's max, so maybe I'd get 1/3 watt on average, so $3/watt. That sounds like a payback period of six years - a no brainer...

Of course, that's without installation costs. Presumably, this kind of thing will make sense for new construction first. Especially in, say, Las Vegas.

Philip Sargent said at July 19, 2005 2:59 AM:

I think that would be $1 per peak Watt (electric), that's the usual way of quoting most renewable electric generation. Over a year, the total number of hours of peak generation will be much less than 1/3 the number of hours in a year (nighttime, low-angle mornings and afternoons, clouds...).

From the original article:
That "39% at 236 suns" means 61% of the energy is heat, and at that intensity it must need some pretty good cooling. 1 sun is about 200/sq.m. (in the UK, peak?), so 61% of 236 suns is 2.9 W/sq.cm. OK, so air cooling can cope with that, but I hope that heat-sink cost is included in the system cost. (A Pentium 4 is 10x more energy intense at about 30 W/sq.cm I think.).

Since you have to pay for the cooling installation anyway, it makes sense to use the heat for something. At 236 suns that could be a very useful temperature (250 - 400 C?) except that the cells presumably have an optimum operating temperature of less than 100 C (?).

So there is a trade-off: it is worth it (in energy use terms) if you can run the cell hotter at a lower PV efficiency if that means that you can get better utilisation of the "waste" heat (e.g. by a Stirling engine generator) - but that would add to the installed cost per peak Watt considerably.

One good use for all that low grade heat would be for sea water desalination - more likely to be useful than domestic space heating in the places that this sort of solar power station would operate.

Philip Sargent said at July 19, 2005 3:16 AM:

That figure of 200 W/sq.m. is a bit low for the best sites. In Australia the peak is above 1000 W/sq.m at Wagga for an average daily insolation of 6.5 kWh (thermal)/day.

So, at 1 $/W (electric) peak at 33%, $1000 will buy you 3 square metres of collector.
3 sq.m. will produce 6.5kWh (electric) over a day (using the same 33% efficiency figure and assumed to be valid for low-light conditions, which is unlikely). Over a year this is nearly 2,400 kWh.

At 6.2 cents (USA) per kWh that is $147. That is a payback period of 6.8 years. Not bad at all.
However at 3$/W peak electric, it is over 20 years still.

Brock said at July 19, 2005 5:33 AM:

"Cheap photovoltaics will give an economic advantage to areas such as Arizona where heavy sunlight falls."

Arizona? How about India, or even Ghana? One of the chief barriers to investment in many developing nations is lack of a reliable (or even existant) electrical infrastructure. If a commercial roof in CA can pay for 90% of their power needs with Sunflowers, I bet a commercial roof in Africa could sell electrical power to its immediate neighbors. Since the Sunflower is extremely cheap (compared to mega-powerplants like nuclear), it could works its way slowly into smaller economies, powering homes & home busienesses.

A C said at July 19, 2005 6:11 AM:

Rooftop solar power systems don't have to compete with utility generation costs. They have to compete with the cost of power delivered to the end user. So in many cases they are already cost-competitive.

Eric H said at July 19, 2005 8:33 AM:

I think Brock and A C are right: solar is already cost-effective when you live in a well-insolated area away from the grid because you don't have to pay for pole and line installation. It is probably cost-effective in a well-insolated area for new construction because it doesn't have to be retrofit. Undeveloped Africa is ripe for this technology. The key is insolation, and most of the American West is above 5.5 hours/day on average (this time is the average amount of energy actually delivered by the sun to a location divided by the amount of power delivered at noon on a cloudless day).

A payoff time of 20 years may sound like a long time, but it can always be incorporated into the sales price of the home. I think the biggest problem is that PV systems are rare and people are likely to discount their utility because they don't have a good feel for lifetime operating costs. In other words, if I buy a house with an installed 10-year-old system, will it still be operational in 10 more years? What about the inverter? What about the grid-tie electronics and islanding? What about grid-tie laws? Dropping the payback time down below 10 years will provide psychological relief so people will be more receptive to buying a house with PV already installed, and that reduces the potentially unrecoverable cost (loss) to someone considering whether or not to retrofit their home with solar.

Anyone want to bet that the subsidy programs will not stop even after the prices fall that low? I personally believe that the subsidies raise the price of PV systems. If someone can sell all the 1 kW systems he can build at $5000, then the government offers a $1000 subsidy to buyers, why wouldn't the seller raise his price to $6000? It would still look like $5000 to buyers, and they will sell just as many, but their per-unit profits go up. I'd guess they would spend at least a little of those profits on politicians.

For people making calculations against their own bills, keep the units in mind. The PV systems are rated in kW at peak *power* production, i.e. how much power they generate at noon on a cloudless day if the panels are aimed at the sun. Your bill is in kW-h, or *energy* consumption. Energy is power expended over time: energy is measured in joules, power in joules/second. Another name for a joule/sec is a Watt. A Watt-hour is 3600 joules (1 W-hr = 60 W-min = 3,600 W-sec = 3,600 j/sec * sec). A kW-h is 3,600,000 joules. To figure your energy generation from a 1 kW (peak) solar system, multiply by the average daily insolation (from the map Randall posted) and by 365 days and call this "Annual energy production". To figure your energy cost from solar, divide the system cost by the annual energy production. The system cost is a combination of the collection system, the electronics, and installation.

Randall Parker said at July 19, 2005 10:53 AM:

Check out these maps of solar insolation in the United States at different times of the year. You can get the same maps closer together on the page. Seattle is probably the worse American city for solar power. No wonder some people who live there get depressed by the grayness of the clouds.

Also, here are some global insolation maps. Note the intensity of the light falling on Antarctica during its summer season when the light falls all day.

Here are some tables on insolation by American city. I think these tables make the differences between the cities seem less dramatic than the maps do.

One thing that strikes me from the global maps is that if solar becomes cheaper than other power sources then it would make sense to move aluminum smelters and other high energy industries to places where the sun is intense all year around. But some of the best places for this are either in chaotic and corrupt Africa or over oceans. The very top of Australia might be a good place to put aluminum plants once solar photovoltaics become cheap.

Randall Parker said at July 19, 2005 10:56 AM:

Eric H,

The houses that operate totally off-grid have to have lots of batteries. Those batteries take up space. they have to be replaced periodically. How often do the batteries have to be replaced and at what cost?

I keep arguing for increased funding of electrochemistry research because better batteries are a key enabling technology for solar (and wind and nuclear). We need better cheaper longer lasting lighter batteries for cars. We also need them for houses.

Rob said at July 19, 2005 11:25 AM:

Only the first problem is technological. The much more difficult problems are in our infrastructure.

Our power grid is not designed to support widespread generation and it will require a huge, costly, complex and - in the beginning at least - somewhat risky upgrade. Think about it: at any instant, the amount of power flowing into the grid must equal what is coming out. Luckily, demand is pretty constant, so power companies can switch plants on and off the grid pretty deliberately. Imagine, however, a southern city of 1,000,000 people that is getting a good percentage of its peak-demand power from rooftop solar arrays just as a big, dark thunderstorm slides in and covers the city - a bunch of electricity suddenly has to come from somewhere else. At a very minimum, power companies will have to be able to monitor and control all generation coming into the grid (and it would sure help if there was some way to efficiently store power in commercial quantities).

Consider this: it took about 60 or 70 years to make telephone access essentially universal. A dozen years after broadband internet access became technologically feasible (and it's very low in capital requirements by comparison), it's just starting to become widespread. We have to have electricians who know about this stuff. We'll need architects to figure it into plans and we'll need builders whose carpenters know how to frame for the extra load. Insurance companies will have to deal with it (am I liable to the power company if my unit fails during a peak-demand period? Is my home insurance bill higher or lower once I have that thing on my roof?) and so will mortgage companies. Home inspectors will have to know how to tell how much life is left in an existing system, just as they do today with roofs.

And so on. It's not that there aren't solutions to all of this, just that the infrastructure solutions seem to be lagging behind the basic technological challege of generating the power in the first place.

My guess is that it will take at least 15-20 years for rooftop solar to become widespread - once it becomes generally cost effective, which I think is still some years away.

Eric H said at July 19, 2005 12:08 PM:


That's a good point about aluminum and other manufacturers, who were among the first to develop hydroelectric power. And yes, lots of batteries; that ends up being the major cost component, IIRC, especially if you look at lifetime costs. Not only do they take up space, but you need to ventilate them because of the H2 buildup during the charge phase. I think you are right about the need to invest in batteries. Luckily, grid-tie allows PV to be the bridging technology to off-grid PV.


Once people have gotten over the bogus "deregulation" of the electric power industry in California, we need to try real deregulation. Power companies will have to start thinking about how they can use solar rooftops rather than fighting them. The grid-tie systems are set up for it (including IEEE-compliant, automatic disconnect during blackouts so you don't end up shocking the line workers or trying to power your neighbor's 10 kW hot tub with your 1kW PV system). When power companies need to expand their capacity, they can subsidize homeowners rather than spending money on lawyers, lobbyists, and senators to fight for new building permits for coal-, gas-, or nucular[TM]- fired plants. Eventually, in a truly deregulated environment, I would expect to see power generators (large-scale and cogeneration), line leasers (who operate lines, switching equipment, dumping equipment, and - eventually - storage equipment), and users trading energy and line capacity. If you don't trust the grid, you will be able to get off of it cheaper than you can now. People are already complying with relatively new standards to support widespread, grid-based generation (google "IEEE 1547" and "IEEE 929"), so it's further along than you think.

Eric H said at July 19, 2005 12:19 PM:

BTW, there is a good workbook at the address below that steps you through the process of designing a PV system (grid-tie or off-grid, including battery sizing):


and information on the latest industry prices is available at http://www.solarbuzz.com/

gmoke said at July 19, 2005 3:29 PM:

What's the payback period on a furnace or an air conditioner?

michael vassar said at July 20, 2005 8:34 AM:

Since Au is a major source of Al ore, making Al in Au would also save shipping costs.

Brock said at July 20, 2005 10:40 AM:


Lots of stuff like AC or furnaces don't have a payback period unless they're used for a commercial purpose. If a restaurant buys a fridge, maybe that produces revenue and savings of $X / year, in which case the payback period can be measured. If I buy an AC unit for my living room, I get a quality-of-life upgrade. That's not as easily measured in dollars though, so to speak of "payback periods" doesn't make any sense.

For consumers, electricial generaion always have a payback period measured by your savings at the meter. That's true for both homes and businesses. That's why it's a useful measure in this case.


Eric H,

No, in this case subsidies do not create disincentives. They are simply the only reason there's a market at all. Technologically, solar just can't compete with fossil fuels & nuclear on a $/Kwh basis.

That being said, I think the subsidies will be phased out as solar uptake continues. It simply wouldn't be feasible to subsidize a major player in the energy market. I think what will eventually replace the subsidies however is a "pollution tax". Solar and wind would be tax free, while nuclear, natural gas, oil and coal would (roughly in that order) face higher amounts of taxation. The reason I say this is because the government picks up the tab for a lot of the externalities produced by pollution. Medicare/ Medicaid pays for the health problems. The EPA has to be funded somehow. Nature site clean-ups that can't be pinned on any one company still need to be cleaned up, etc. A pollution tax would efficiently price these costs into the cost of electricity generation, and create incentives for pollution-free energy such as solar.

Randall Parker said at July 20, 2005 11:48 AM:


If a newer furnace or air conditioner replaces and older and less efficient one then there is a payback period due to decreased electric power usage.

As for subsidies: They are increasing. Congress just voted up higher tax subsidies for wind and solar. Unfortunately the subsidies really distort the market because passive designs, direct solar (e.g. heating systems), and better insulation do not get equal tax treatment. I've posted here before comparing much shorter payback periods for more passive solar designs for water and space heating as compared to photovoltaic-based installations. The subsidies create disincentives for non-photovoltaic solar and insulation and passive designs. I see this as a serious error.

As for taxes: The nuclear power industry already is taxed to fund Yucca Mountain and associated programs for nuclear waste disposal. The fund set aside for this purpose has tens of billions sitting in it unspent because of political opposition to a centralised waste disposal facility. So I don't see how nuclear currently is producing untaxed external costs.

Nick said at July 20, 2005 1:09 PM:


Here's an interesting conclusion from a IEA report (http://www.oja-services.nl/iea-pvps/ar01/task5.htm):

"Factors to decide on the limit of PV system penetration in distribution networks were listed and an analysis based on distribution line voltage limit was conducted. Considering the distribution system line voltage control operation, there will be no problem if the PV systems are penetrated below minimum load level (typically 20% to 25% of distribution line capacity). Measures to stretch the limit of PV penetration and some financial aspects of PV penetration were also discussed. Distribution system planning considering PV system penetration and demand side management using PV system will be the future key to enhance PV penetration."

So PV capacity of up to 20% of grid capacity seems to be no problem, with current systems. Similar to wind, which kind of makes sense.

gmoke said at July 20, 2005 9:10 PM:

"If I buy an AC unit for my living room, I get a quality-of-life upgrade."

I believe there are many people who go solar as a quality-of-life upgrade without thinking about payback. I would venture to say that some people buy solar walk lights who would never think of putting in another kind of outdoor lighting.

IMHO, solar development starts from low voltage DC applications using rechargeable AA and other tradable batteries and proceeds up to 12 volt systems to interface with cars and other vehicles along with the parallel development of whole house and industrial PV systems. With low voltage systems, PV solar is already cost effective if battery switching using reliable rechargeables is available. These markets are in emergency, camping, and individual systems - light, a radio, a phone, maybe a computer.

What's the payback on one of those solar backpacks? Immediate, if it's a quality-of-life upgrade. What's the payback on the white LED solar light clipped to my own grotty backpack? It cost $30 and is competitive within its market in its use as a bike light.

The base assumption of the usual Futurepundit discussion of PV solar seems to be that PV solar is only cost-effective when one roof can provide the current family electrical demand without any conservation, efficiency, or further modification. I think that's not the best way to think about it. People who buy solar are also probably going to be more energy conservative and efficient than those who don't. Certainly, the early adopters tend to be more energy conscious. They are going to try to reduce their load as much as possible before investing in the proper scale of solar PV. The electrical load of most solar houses may be lower, perhaps much lower, than today's average. The objective is to get to solar products and services that equal or surpass the ones we have today at less long-term cost, solar products so easy to use they become unconscious, "normal."

What might a state-of-the-art electrical load look like? I use about 100 kwh per month to provide light, refrigeration, TV, computer, printer, fans... I believe that's on the low end of the scale. Can it be lower? How much PV would that require here in the Northeast? How much might it cost? Would a one window system be more affordable as well as marketable for multiple use (camping, security in case of emergency or disaster...)?

The Futurepunditocracy seems to focus only on shingling all our roofs with PV solar. I think that we're gonna see things like a truly decent solar rechargeable flashlight before that happens because PV is affordable in the flashlight marketplace. If a small scale mass market PV solar product does break through, then the likelihood of PV shingling everything will be all the greater.

Randall Parker said at July 20, 2005 9:50 PM:


Some people do put up solar walk lights and then discover that they fail after a few years. I wonder if one of the brands last longer. But at one family member's house they were neat while they lasted. I'm guessing high moisture air (in Forida) caused then to fail.

You state:

The base assumption of the usual Futurepundit discussion of PV solar seems to be that PV solar is only cost-effective when one roof can provide the current family electrical demand without any conservation, efficiency, or further modification.

No, the assumption is that PV solar will be cost effective when it costs a lot less. You can spin it any way you want but for the vast bulk of users it costs too much. Its usefulness for some niche purposes isn't going to do much to decrease fossil fuels consumption. PV solar isn't going to make major inroads into fossil fuels use until it costs a lot less. Spinning and praising it or praising those few who use it will not change this basic fact.

gmoke said at July 21, 2005 2:56 PM:

Please tell me what you think is "praise" and "spin" in the post in question.

Jim said at July 22, 2005 9:21 AM:

RP - I wonder if you caught the article about nuclear power in the economist? I read the print magazine, but the online is paid subscription access only. Anyhow, they had an assessment of the relative costs of gas, coal, nuclear, and wind. They are all same order of magnitude costs, but what was really intersting to me was the breakdown of costs between capital costs and fuel costs. Nat. gas has by far the cheapest capital costs but fuel costs are the highest (and also the biggest variation), while wind and nuclear costs were mostly capital, as one would guess.
What occured to me is that in the short to medium term (before better batteries), natural gas plants can play the role of energy storage in the sense that can easily come online to compliment fluctuations in Pv or wind power. I could imagine an optimum utility energy blend of something like 70% baseline coal/nuclear 30% wind/PV and 20% natural gas/batteries. (yes that adds up to 120% intentionally) and hydro has best of all worlds, where available.

Randall Parker said at July 22, 2005 10:00 AM:


If you would avoid complaining about others and focus more on providing facts and logic you'd get into fewer pissing contests and discussions would be more productive.

Randall Parker said at July 22, 2005 10:08 AM:


No, didn't see The Economist article. I let my subscription lapse years ago.

I am already aware of the relative capital costs and capital to fuel ratios. Natural gas has become uncompetitive as natural gas fuel costs have risen so much in recent years. This is why so many utilities are planning to build coal burning plants. Higher capital costs than natural gas and more polluting. But the fuel is now cheaper and plenty of coal still sits in the ground, enough for many decades to come.

Natural gas is already heavily used for peak demand. Add in solar and wind and then natural gas would be needed to fill in when the sun doesn't shine and the wind doesn't blow. Even if solar and wind electric become as cheap as other forms the fact is that they are not 24x7. So money would still have to be spent on capital to make the gas burning plants (or coal or nuclear) for when the wind doesn't blow and the sun doesn't shine.

BTW, I'd rather not have 70% baseline coal. I want cleaner air and until the coal burners accept tougher regulation of their emissions I'm opposed to coal. But if they were to accept the tougher emissions regulations then coal would probably become too expensive. Coal might even become more expensive than wind. But I can't find info on what truly clean coal would cost.

Jim said at July 22, 2005 11:48 AM:

RP - I'll see if i still have the article when i go home, and just type in the costs that they quote. I know these are figures that you've looked for data on before.
and yeah, 70% coal would be lousy. my preference would be heavy on the nuke, light on the coal to make up this arbitrary 70% fixed work-horse power number.

gmoke said at July 22, 2005 4:08 PM:

Mr Parker, I ask you to point out exactly what my "spin" and "praise" consist of and you respond by accusing me of instigating a pissing contest and complaining about others. This leads me to believe that you can't explain yourself. Too bad.

I think the "deus ex machina" and "magic bullet" mindset is basically juvenile. I like technology at least as much as the next guy but technology ain't gonna save us from our previous misuse of technology. Imagination and innovative organization may do so but I don't believe any black box is gonna make it Christmas every day. Futurepunditocracy seems to like large scale tech and sweeping solutions. I tend to like appropriate tech and incremental steps. If you, Mr Parker, can't deal with that, tough. That ain't spin or complaining, that's a different point of view. It's what debate and the free exchange of ideas is supposed to be about, or so I thought. Obviously, you think differently.

Pen said at July 24, 2005 11:56 PM:

To figure your energy generation from a 1 kW (peak) solar system, multiply by the average daily insolation (from the map Randall posted) and by 365 days and call this "Annual energy production". To figure your energy cost from solar, divide the system cost by the annual energy production. The system cost is a combination of the collection system, the electronics, and installation.

Jim said at July 25, 2005 7:00 PM:

from economist 7/9/05 p. 59, chart:

mid-term generating costs of new power plants, $/MWh

combined-cycle gas turbine:~31 to 40 total= ~9 to 29 (fuel) + ~3 (operation&maintainence) + 7 (investment)

coal:~34 to 42 total= ~11 to 20 (fuel) + ~5 (operation&maintainence) + 17(investment)

nuclear:~45 total=~5(fuel) + ~14(operation&maintainence) +~26(investment)

wind:~43 total=2(operation&maintainence) + 41 (investment)

a 5% discount rate is assumed for investment $. it assumes good wind sites. it's referenced from the International Energy Agency (IEA) World Energy Outlook 2004

The article is overall positive about the growing political coalition for nuclear, particularly in n. europe where it's currently banned. the french claim that the standardized plants they use cost 20% less to operate and 30-40% less for capital costs compared to Britain where it's mostly one-of-a-kind plants.
It is a bit skeptical of the economics of nuclear though, pointing out that it's dependent upon $6 Mbtu gas prices rather than $3.50 not so long ago. It is doubtful about the nuclear industry promises of $1,500 to $2000 per kW of installed capacity. At this price, the economist is basically recommending coal, while noting the lack of CO2 emmissions is beneficial. They end by noting the convergence of Bush and environmentalist in U.S. for nuclear.

Randall Parker said at July 25, 2005 7:34 PM:


Some big US electric powerplant operators think nuclear is now less expensive and want to build nuclear plants. I came across a quote from one big electric plant operator exec to this effect a few months ago. Can't remember if I blogged on it. But this is consistent with the desire of several power plant operators to get the government to agree to subsidize the regulatory approval process for the first construction of each of 3 new nuclear plant designs (GE, Westinghouse and I think Areva or something like that).

Also, I am skeptical of the natural gas versus coal difference. The electric plant operators shifting toward coal in a really big way. They must think coal is cheaper. I figure they know better than The Economist. Ditto on wind. We'd see a lot more wind site construction if it was that competitive.

I also wonder what level of cleanliness is assumed for coal.

Also, location matters. Coal is cheaper near coal fields. Wind doesn't work for the US southeast in particular and for some other parts of the country.

Jim said at July 25, 2005 11:14 PM:

Overall the tone of the article made it sound like there was a pretty good chance of it happening in the U.S. soon, partly due to the politics - you better have support and as much pork as possible before setting up these mega-deals - the capital costs are higher than coal or nat. gas. The economics call for a very large time and capital commitment, with modest mean and low variance fuel costs. Catastrophic risk is the biggie that when solved through gov't gift will kick the domestics into gear.

The economist author made the point that coal is the most logical/economical. Not only is it the cheapest, the enormous known reserves make the price very stable and predictable - exactly the kind of thing you're looking for if you're a utility investor. no mention of coal cleanliness.
Nat. gas is the cheapest mean, and substantially cheapest capital costs, but the variation on the fuel costs is huge and unpredictable.
Wind is barely mentioned, and is really a niche source that only makes sense in wide open, sparse and windy land especially near big transmission or a flexible industrial consumer. The cost is almost all upfront capital, but it can be incremental, and is a bit low considering they assume 5% discount rate - good luck finding a bank to do that one. It does have the advantages of no fuel, lowest maintainence costs, compatible with farming.

Wind and natural gas seem to be complimentary in a number of ways (particularly if you're in the great plains) such as hedging predictibility of fuel costs with predictability of generation capacity.... especially if you could generate nat. gas somehow with excess turbine power - i suppose you could make H2 by hydrolysis and burn that in your turbine that was specially designed. power would probably be double the price, but clean and reliable (again if you live in fargo, etc.)

Nick said at July 26, 2005 12:17 PM:

*Ditto on wind. We'd see a lot more wind site construction if it was that competitive.*

It's my understanding that currently wind construction is constrained by turbine supply. I think you have to keep that in mind, when evaluating it's market share.

Jim, I'm struck by fact that the Economist projects that wind is less expensive than nuclear, and only $1-3/MWH more than the upper range of gas and coal. I suspect you're being a bit conservative on wind - for instance, West Texas is pretty close to the large California electric market, and there's substantial midwest potential near Chicago. Would you agree that wind can supply up to 20% of overall KWH's with just the current technologies for supply and demand management?

Randall Parker said at July 26, 2005 12:48 PM:


Where'd you read that turbine supply is a bottleneck?

I would not, in any case, expect long term project planning to be limited by present supplies. Manufacturers can ramp up production for just about anything pretty quickly. Capital equipment construction lead times are not that long. So I'd expect to see a lot more wind site proposals by electric generator plant operators.

Right now in the United States about 100 coal fired plants (and these tend to be in the range of several hundred megawatts and higher for each site) are at various stages of planning or early construction. If wind was truly competitive I'd expect to see something similar with wind sites. So where is the planned construction for tens of gigawatts of wind sites?

Greg Watson said at July 27, 2005 12:23 AM:

Hi Guys,

Just a few comments from down under.

My company, Green and Gold Energy, has developed a very innovative rooftop solar PV concentrating tracker called the SunBall(tm).
More info available here:


A 4 SunBall(tm) system will cost about $4,400(US$)(1.5kW inverter extra) with ~1.4kWp, ~3,400 kWh annual generation and 25+ years of life. That calcs kWhs at ~$0.0517 without any government assistance. With the advantage of Max kWh Tracking(tm) and "Cool PV"(tm) cell temp, very significant generation is available on hot summer afternoons when flat panels die a heat death.

Comments most welcome,

Nick said at August 4, 2005 2:04 PM:


I can't find my source on turbine supply backlogs - the best I could find was info that GE's wind revenues doubled year over year. However...what I read about a month ago was that all scheduled production of wind turbines, at least for GE, was promised to specific customers through the end of 2005, which was the expiration of the credit, at that time. This meant that, combined with installation delays, that planning for new projects was delayed because no new projects could be scheduled to be completed before the expiration of the credit.

I think there are several relevant factors to the question of planned construction of wind.

First, compared to coal it's starting from a small base, so it's not going to be overwhelmingly visible, on the other hand it's big enough that it's not going to expand by more than, say 75% per year, because of the logistics involved in expanding production. Good examples are solar, which is constrained by silicon production, and hybrid production, which is constrained by battery availability (per Ford, which says the Escape's production was limited to 20,000 per year by battery production).

2nd, it's new. I think a lot of utilities don't really trust it yet, or believe that natural gas prices (and coal for that matter, which I believe is up 25% lately) are going to stay high. They just don't have a lot of experience with it, or the logistics of dealing with it's intermittency. Look at GM and Daimler/Chrysler: you would think they would have done hybrids long ago, but they just didn't trust that hybrids weren't an expensive fad. Toyota took a longer view (aided by better cash flow, it must be admitted), and won the bet. It just takes a while for people to get used to new things. Plus, there's infrastructure to be developed: financing, for instance, which is complicated by the need for off-setting profits for the tax credit.

3rd, the tax credit's on-again, off-again status has screwed things up. In 2004 everything went on hold, waiting for the tax credit to come back. Same thing in prior years. Makes it hard for anyone (especially suppliers who would have to spend big on capital to expand) to plan, or trust in demand. This is true even for those projects that could be justified without the credit - why build now and make 10% ROI, when you can wait a year and make 25% (remember, the credit all depends on when you finish building - finish building in a year when it's not in force, and you never get it, for the life of the project)?

Greg Watson said at August 17, 2005 8:27 PM:

Hi All,

SunBall web site now operational:

All the best,

Harry Houck said at November 4, 2005 11:01 AM:

Payback, schmayback! What's the payback if you need to remove lead paint from your kid's room? Instead of asking the payback for PV, ask what's more expensive, fighting over vestiges of non-renewable oil or converting sunlight to electricity? We need to stop burning oil.


Post a comment
Name (not anon or anonymous):
Email Address:
Remember info?

Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright